run-3/checkpoint-500/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.5723270440251573,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7209677419354839,
      "eval_loss": 3.2341866493225098,
      "eval_runtime": 2.427,
      "eval_samples_per_second": 1277.299,
      "eval_steps_per_second": 26.782,
      "step": 318
    },
    {
      "epoch": 1.57,
      "learning_rate": 1.4758909853249476e-05,
      "loss": 3.7475,
      "step": 500
    }
  ],
  "logging_steps": 500,
  "max_steps": 1908,
  "num_train_epochs": 6,
  "save_steps": 500,
  "total_flos": 130007874759588.0,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.14437383131434334,
    "num_train_epochs": 6,
    "temperature": 16
  }
}